Custom Query (3220 matches)


Show under each result:

Results (28 - 30 of 3220)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Ticket Resolution Summary Owner Reporter
#35 fixed Finding things in the D1 array jeff s.keeley@…

Hi, Lois and I had discussed finding and writing things to the UM in the D1 array and not really knowing how to go about it.

Whilst reading some of the umui help pages I stumbled across this which suggests that fields do get written to the D1 array and there is a way of finding out what the pointers are using UM routine FINDPTR. I was wondering if anyone knows where this routine might be? and if it is part of the standard UM setup….

Thanks, Sarah

Help for window : atmos_STASH_Usage

This window allows users to define the attributes of the usage profile within the STASH windows.

The useage profile defines the output destination of a diagnostic.

Most often, simple diagnostics are written to one of the user-PPfiles on units 60-69. The user can specify these as files to be reinitialised or recycled (that is, new files are attached periodically to this output stream to avoid the files becoming too large). In this case, each stream should be dedicated to a particular sub-model. Several streams can be attached to one sub-model. When the streams are defined as not reinitialised, users can mix output onto these streams as they please. See subindep_PostProc_PPInit.

Alternatively, diagnostics can be written to the D1 area (and the external dump by implication). Users may wish to do this so that diagnostics can be passed from one section of code to another. This method is used by several of the STASHmacros. To allow the code to find the particular diagnostic it is also Tagged and search routines can find the STASHlist record that has such TAGs attached. See UM routine FINDPTR. These are the TAGs used by various code sections:

10 atmosphere to ocean coupling 11 ocean to atmosphere coupling 20 atmosphere to slab coupling 21 slab to atmosphere coupling 30 atmospheric assimilation dump extension

If the variable is not required in the dump (that is it is not time processed or you are certain that all time processing occurs within the dump intervals) you can similarly write tagged diagnostics to secondary store. These are not written to the dump as they are in an extended area of D1.

Alternatively, diagnostics can be written to the D1 area (and the external dump by implication) and tagged to be picked up by the climate meaning system. Users should attach this sort of profile to diagnostics from sections other than the climate mean sections (21-24). Users may wish to perform time processing (such as daily means sampling values every timestep) and have these further processed by the climate meaning system. This is the only way to obtain climate means with values accurate to the nearest timestep. Such diagnostics need to be written to the climate mean files for each meaning period. Specify the periods required, the tagging will be added automatically from this information.

Diagnostics from the climate mean sections (21-24) are either "mean prognostic variables" or diagnostics calculated from those variables. They are the result of the climate mean dumps. The sampling frequency for such meaning is the basic dumping frequency of the run for a given sub-model. Such diagnostics are written direct to the climate mean files.

Dr Sarah Keeley Climatic Research Unit University of East Anglia Norwich NR4 7TJ UK

tel: +44 (0) 1603 592721


Dr Sarah Keeley Climatic Research Unit University of East Anglia Norwich NR4 7TJ UK

tel: +44 (0) 1603 592721


#37 fixed W field in stash that works/UMUI windows jeff caroline@…

Hello NCAS,

My last help ticket got deleted so I'll try again!

Can you recommend a W on pressure levels stash please? I tried 15242 but nothing outputted. Am I missing another one?

Also, here at Leeds we have the problem (or I always do anyway), that the umui windows often outsize the screen. Normally I can get round this by moving the window within the screen but it's a problem for the load diagnostics page because the window is too long to be moved within the screen so I can never see the bottom of it! This might be the reason why I have missed a really obvious W stash. Jason knew about the problem but didn't know a solution and now he's left of course. Is there anything you can advise me to do?

Best Wishes, Caroline

#38 fixed Looking at .leave file jeff caroline@…


I have a basic 6.1 op model (hi-res) and we are trying to reconfigure with the ancil files in /hpcx/home/n02/n02/umx/vn6.1/ancil/atmos/n320 specifically qrparm.soil_igbp_sahara This was just a basic run to check that the reconfiguration works, we have made no other changes to a job that I know works (xaexc taken from xaexa). However no xaexc.astart file was produced so I know there is a problem somewhere.

Me and Chris tried to look at the .leave file in my umui_out folder but it is huge and I can only see to halfway down so can't see the error output at the end.

Everything with the job seems like it should work, so it isn't obvious why the leave file should be so messy.

Let me know if you need more info, I've changed permissions in my files on hpcx so you should be able to see everything.

Many thanks, Caroline

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Note: See TracQuery for help on using queries.